It has started: Q6600 for 245 Euro and still going down

[jspace]

^^ Agreed. I LOVE AMD but right now it would be dumb to get one unless you're going for low power.

[Vapor]

I'm still not convinced with G0s....and I LOVE B3s (then again, I've been lucky ).

mwave deal sure does look good....

[[XC] serlv]

Agreed SB

I'm sure if in the future other CPU's prove to be better value for crunching then we will buy them too

Hell yeah! We like good crunchers.

Had an army of XP and XP-M's. Those things were crunching beasts.
Had a small fleet of AMD Athlon64 type, mainly DCs, both X2 and Opteron.
Have a small squad of C2Ds and a couple of Octohamsters.
Gonna get a quad or two, once the dust settles.

If AMD jumps back in the fray and puts out a stunner, I'll get that, too.
I certainly have nothing against AMD. I just want the most bang for the buck.


Whatever works the best is what I get.

[kiwi]

Guys, not all G0 are good!

[SiGfever]

Check out the Quad...

http://www.newegg.com/Store/SubCateg...PUs-Processors

[sierra_bound]

Hmmm, Newegg is already sold out of Q6600's at $300. But Mwave still has them, though it raised the price a little to $278. Zipzoomfly has the Q6600 for $295. Fry's still wants $565.

[Vapor]

Guys, not all G0 are good!

There's a lot about G0 I don't like....and yes, you're absolutely right.

Basically, if I were in the market for a CPU right now, I'd DEFINITELY get one of those mwave B3s.

[sierra_bound]

I think until people actually start testing the retail G0 quad-cores, it's too soon to reach a verdict. The price drop took effect on a weekend, so most of the CPU's won't be shipped until Monday. A lot will depend on how these chips were binned. The G0 E6850, for example, seems to perform a lot better than the other G0's with lower multipliers.

[Brother Esau]

I have a fella that is buying my E6420 and pushing for it pronto so I went and ordered a Q6600 from mwave this evening I am thinking it will be G0 Stepping personally. Any comments on the B3 Stepping in case it ends up being that?

I still have my Q6600 G0 order pending with Tank Guys in que at position 204

[sierra_bound]

I don't know how people can feel disappointment when they paid only $270+ for a CPU with four cores. Of course, everyone would like to pay next to nothing and get a chip that's a real screamer. Even at 3.1-3.2GHz, Kentsfields are very responsive. I doubt many apps are going to struggle at that speed.

[[XC] DragonOrta]

There's a lot about G0 I don't like....and yes, you're absolutely right.

Basically, if I were in the market for a CPU right now, I'd DEFINITELY get one of those mwave B3s.

What about the G0 is it that you don't like?

[Brother Esau]

LOL...whats not to like about a quad core at those prices? I am interested in lowest temps B3 G0 I could care less!

[Vapor]

What about the G0 is it that you don't like?

For the QX6850 and the E6850, I think they're great....BUT, in every other model, there are more than a few things to worry about.

1) Some have nasty FSB walls.
2) Many have FSB issues when not used on P35 chipsets (which I never plan to run). X38 will probably be okay as well.
3) Coldbug and LN2 performance seem to be an issue already (not applicable to DC, but it's a worth a mention).
4) It's a budgeting revision....that's HUGE.
4a) it means there's a very high likelihood of future G0s being piss-poor clockers (at least they'll run cool)
4b) it means that yields are higher, at the expense of consistent clocking chips (i.e., there will be more duds, thus more complaints)


And for a better-versed version of what I think: http://www.xtremesystems.org/forums/...8&postcount=85 (and he's not a nobody ).

5) my luck with B3s is...very good

[sierra_bound]

I think we're talking about two things here - benching stable versus folding stable. Most of the people here in the WCG section are getting quad-cores because they can increase their production at a very reasonable price. The difference between crunching with a quad-core at 3.2GHz versus 3.4-3.5GHz is not really huge. Yes, you will get more daily points at the higher speeds. But is it enough to justify paying $1000 for a QX6850 versus $270 for a Q6600? For most folders, I doubt it.

[[XC] riptide]

For the QX6850 and the E6850, I think they're great....BUT, in every other model, there are more than a few things to worry about.

1) Some have nasty FSB walls.
2) Many have FSB issues when not used on P35 chipsets (which I never plan to run). X38 will probably be okay as well.
3) Coldbug and LN2 performance seem to be an issue already (not applicable to DC, but it's a worth a mention).
4) It's a budgeting revision....that's HUGE.
4a) it means there's a very high likelihood of future G0s being piss-poor clockers (at least they'll run cool)
4b) it means that yields are higher, at the expense of consistent clocking chips (i.e., there will be more duds, thus more complaints)


And for a better-versed version of what I think: http://www.xtremesystems.org/forums/...8&postcount=85 (and he's not a nobody ).

5) my luck with B3s is...very good

Yep. It almost is a small repetition of when the new >31 week conroes came out. The Jury was out as they failed to be ACROSS the board great clockers.

[Vapor]

I think we're talking about two things here - benching stable versus folding stable. Most of the people here in the WCG section are getting quad-cores because they can increase their production at a very reasonable price. The difference between crunching with a quad-core at 3.2GHz versus 3.4-3.5GHz is not really huge. Yes, you will get more daily points at the higher speeds. But is it enough to justify paying $1000 for a QX6850 versus $270 for a Q6600? For most folders, I doubt it.

I'm just saying G0 isn't all it's cracked up to be, especially at the $270 price point.

The difference between 3.2 and 3.4-3.5 is 5-9%....that's fairly significant, IMO. This isn't QX6700 vs. Q6600 here, it's G0 vs. B3. Yes, B3 is hotter, but at the low end (and Q6600 is somewhat low-end), G0 will suffer and could be even worse than the 5-9% disadvantage when you average it all out.

I guess I'm being overly negative about G0 to offset the excitement and bring it back down to earth....but give me a choice between a random G0 and a random B3, and I'll take the B3. This is XS, we don't stick to stock HSFs where a G0 will typically be a little better, in fact, we use the best HSFs....

Oh, and voltage response is another concern of mine on G0.....especially for 24/7 stability. (but that's largely unfounded so far, but we'll see)

EDIT: and I apologize for the negativity.....taking out my WC-section frustrations here is not my intention

[Movieman]

I'm just saying G0 isn't all it's cracked up to be, especially at the $270 price point.

The difference between 3.2 and 3.4-3.5 is 5-9%....that's fairly significant, IMO. This isn't QX6700 vs. Q6600 here, it's G0 vs. B3. Yes, B3 is hotter, but at the low end (and Q6600 is somewhat low-end), G0 will suffer and could be even worse than the 5-9% disadvantage when you average it all out.

I guess I'm being overly negative about G0 to offset the excitement and bring it back down to earth....but give me a choice between a random G0 and a random B3, and I'll take the B3. This is XS, we don't stick to stock HSFs where a G0 will typically be a little better, in fact, we use the best HSFs....

Oh, and voltage response is another concern of mine on G0.....especially for 24/7 stability. (but that's largely unfounded so far, but we'll see)

EDIT: and I apologize for the negativity.....taking out my WC-section frustrations here is not my intention

Interesting..I think most people are seeing the G0's as dead equal to the B3's but just able to clock higher due to the lower heat output.
If thats a bad assumption we need to know.
I've read a lot of the threads on the G0's and yet to see a negative.
Granted, those are all ES's vs retail.

[Paladin]

And for a better-versed version of what I think: http://www.xtremesystems.org/forums/...8&postcount=85 (and he's not a nobody ).

Since there are really just two bins, locked (Q6600) & unlocked (QX6700), I don't think that quality binning will have as much effect in the quad line as it might in the dual-core. The post linked seems to corroborate this.

[sierra_bound]

I'm just saying G0 isn't all it's cracked up to be, especially at the $270 price point.

The difference between 3.2 and 3.4-3.5 is 5-9%....that's fairly significant, IMO. This isn't QX6700 vs. Q6600 here, it's G0 vs. B3. Yes, B3 is hotter, but at the low end (and Q6600 is somewhat low-end), G0 will suffer and could be even worse than the 5-9% disadvantage when you average it all out.

I guess I'm being overly negative about G0 to offset the excitement and bring it back down to earth....but give me a choice between a random G0 and a random B3, and I'll take the B3. This is XS, we don't stick to stock HSFs where a G0 will typically be a little better, in fact, we use the best HSFs....

Oh, and voltage response is another concern of mine on G0.....especially for 24/7 stability. (but that's largely unfounded so far, but we'll see)

EDIT: and I apologize for the negativity.....taking out my WC-section frustrations here is not my intention

No need for the apology. I guess I'm looking at this purely from a crunching perspective. For the price of one QX6850 you can now get four Q6600's. Even four so-so clocking quad-cores will annihilate one super-clocking QX6850 in daily point production. For those people with large to medium-sized folding farms, the QX6850 does not make a lot of economic sense.

Regarding the 5-9% performance difference, I think it's closer to 5%. But again, is that 5% worth spending an extra $730? For someone who benches and has a lot of cash, it may be. But for the folder on a budget, it probably isn't.

I just noticed Fry's still has not lowered the price on the Q6600. It's still listed at $565.

[Vapor]

S_B, you're still stuck on Q6600 vs. QX6xx0

I'm just resenting G0's existence

FUGGER may have a gem of one, and no doubt they'll probably be the WR chips and the highest-clocking 24/7 chips. But they'll also be the lowest, and probably even in staggering numbers. WCG team will have a large quantity of these (I'd bet), and they'll just not be that good...and on a large scale of the WCG team, that's A LOT of lost points. I'd have rather seen B3s stick around a little longer for like $5 extra

Heck, I know of a G0 QX6850 that can't do FUGGER's single stage clocks....on LN2 (and no, the pot is not at all to blame). Not even the high-end chips are all good.

[sierra_bound]

No, I'm just saying that benchers and folders view things differently.

I can't imagine the obsession with G0's will last very long anyway. We're already into Q3 and Penryn is just around the corner. There's also Barcelona for all the AMD folks.

Once upon a time when I did a lot of benching, I used to think the same way you do. But my views on a lot of things have changed. I don't have a burning desire anymore to have the absolute best chip available. Folding is more about dedication and patience, with a little fun thrown into the mix. It's not about the BOINC CPU benchmark which I think is bogus to begin with. Even a so-so chip will generate a lot of points and complete a lot of work units over time.

[Gautam]

I've read a lot of the threads on the G0's and yet to see a negative.

Yeah, that's the problem.

I was thinking of cracking my knuckles and making a few posts to put things in perspective, but I see that Eric's doing just that.

What's funny is, there's only one retailer I know of even guaranteeing G0's so I'm not sure why its become such a hot topic. Who knows when they'll get them in as so far we have yet to see a single Q6600 G0 retail, and a very small amout of ESes. Who's to say you can get one at this point in time even if you really wanted to?

[Jaco]

cheapest quad i've seen : X3210 220€

http://www.komplett.be/k/ki.aspx?sku=330170